The goal of this fellowship is to provide me with the skills and experience needed to be an independent researcher in the field of substance abuse and social epidemiology. This will be accomplished through coursework, mentorship by my sponsor, Dr. Patricia Kissinger, and co-sponsor, Dr. Richard Scribner, and through my own research project. The objective of this project is to investigate the sexual and drug networks of Latino migrant men in order to better understand the risk factors for the transmission of HIV and other sexual transmitted infections (STIs). This project addresses NIDA's research objective to prevent HIV infection and spread among drug users and their partners. A person's social network is crucial in understanding how their environment impacts their risk of HIV. Within networks, it is important to identify core groups since core group members sustain the disease in endemic periods and promote expansion in epidemic periods. While social network analysis is a powerful tool to identify core groups, it is challenging to carry out. Previous studies have used spatial analysis to identify core groups, however these methods have not been validated. If validated, this would provide epidemiologists with a much easier method to identify high risk core groups. This project has three aims: (1) to use spatial analysis to identify core groups of high risk sex among drug users and non-drug users, (2) to compare the sexual networks of drug using and non-drug using Latino migrants, identify core groups of high risk sex and validate spatial analytic techniques used in aim 1 to identify core groups, and (3) to identify risk factors for belonging to a core group among drug users and non-drug users. My hypothesis is that I will find residual spatial autocorrelation among drug users' sexual networks that is indicative of the presence of core groups. I believe that once I identify core groups using social network analysis in aim 2, membership in a core group will explain the residual spatial autocorrelation seen in aim 1, thus validating spatial techniques to identify core groups. I expect that drug users are at a higher risk of belonging to a core group and that membership to a club or organization will be protective. This study will be carried out by selecting 25 drug using and 25 non-drug using Latino migrant men from a parent study. These men, referred to as egos, will be asked to list and describe all people in their sexual and drug networks. The egos will give referral cards to the people listed in their risk networks, called alters. The first wave of alters, those listed by egos, will be interviewed in the same manner as the egos, and they will also be given referral cards for people in their risk networks. After completing data collection I will have data on the egos, two waves of alters, and information about the second wave alters' networks. Based on data from our parent study, a cohort of 125 Latino migrant workers, we estimate an average of 2.5 people will be reported in a given person's risk networks, totaling approximately 650 people. Data analysis will be carried out in a variety of packages including: SAS, UCInet, and ArcGIS.